GBase 8s JDBC Batch Insert: Configuration and Example

Inserting large volumes of data into GBase 8s can be slow if you execute each row individually. The JDBC batch mechanism — with the IFX_USEPUT=1 parameter — slashes network round trips and boosts throughput dramatically in a gbase database.

Prerequisite

You must explicitly enable batch insert by adding IFX_USEPUT=1 to the JDBC URL. Without it, batching won't take effect.

Full Example

import java.sql.*;

public class Insert {
    public static void main(String[] args) throws Exception {
        String url = "jdbc:gbasedbt-sqli://192.168.50.150:9088/testdb:GBASEDBTSERVER=gbaseserver;IFX_USEPUT=1";
        String username = "username";
        String password = "password";

        Class.forName("com.gbasedbt.jdbc.Driver");
        Connection conn = DriverManager.getConnection(url, username, password);

        PreparedStatement ps = conn.prepareStatement("INSERT INTO t1 VALUES (?)");
        for (int i = 0; i < 10000; i++) {
            ps.setString(1, "2024-10-16");
            ps.addBatch();
        }
        ps.executeBatch();

        ps.close();
        conn.close();
    }
}

How It Works

1. URL Configuration
   
   Append IFX_USEPUT=1 to the connection string. This property is specific to GBASE's driver and instructs the driver to use the optimized bulk‑insert protocol.

2. Batch Execution

   • addBatch() queues a parameter set without sending anything over the network.
   • executeBatch() transmits all queued rows as a single block, so 10,000 inserts become just one round trip.

For even larger jobs, consider splitting input into chunks to keep transactions small and avoid excessive resource usage. This simple switch can turn a sluggish data load into a fast, efficient pipeline on your GBASE GBase 8s database.